Journal of Clinical InvestigaltonVol. 43, No. 10, 1964

The Induction of Proteolysis in Purulent Sputum by Iodides *

JACK LIEBERMANAND NATHANIEL B. KURNICK

(From the Veterans Administration Hospital, Long Beach, and the Department of Medicine,the University of California Center for Medical Sciences at Los Angeles, Calif.)

Iodide, in the form of potassium or sodium io-dide, is widely used as an expectorant for patientswith viscid sputum. The iodides are thought toact by increasing the volume of aqueous secretionsfrom bronchial glands (1). This mechanism forthe effect of iodides probably plays an importantrole, but an additional mechanism whereby iodidesmay act to thin viscid respiratory secretions wasdemonstrated during a study of the proteolyticenzyme systems of purulent sputum.

Purulent sputum contains a number of proteo-lytic enzymes, probably derived from leukocytes(2-4). These proteases, however, are ineffectivein causing hydrolysis of the native protein in puru-lent sputum and do not appear to contribute sig-nificantly to the spontaneous liquefaction of thesepurulent secretions. On the other hand, these in-trinsic proteases are capable of causing proteolysisafter the exudative protein is separated fromdeoxyribonucleic acid (DNA) by the action ofdeoxyribonuclease (DNase) or high concentra-tions of sodium chloride (5).

During these studies an attempt was made toevaluate the effectiveness of various salts for in-ducing proteolysis in purulent sputum, ostensiblyby liberating protein from nucleoprotein com-plexes. Sodium iodide was found to induce morerapid proteolysis than any other salt tested, butwe were able to demonstrate that the mechanismdid not primarily involve the dissolution of nu-cleoprotein complexes. In addition, iodinated ty-rosine and thyronine compounds were found to

*Submitted for publication October 9, 1963; acceptedJune 15, 1964.

Presented in part before the Western Section of theAmerican- Federation for Clinical Research, Carmel,Calif., January 30, 1963.

Aided by a grant from the California Research andMedical Education Fund of the Tuberculosis & HealthAssociation of California and grants HE1880 andCA-0382-07 from the National Heart and National CancerInstitutes, U. S. Public Health Service.

have even greater proteolysis-inducing activitiesthan inorganic iodide.

Methods

Source of sputum specimens. Purulent sputa were ob-tained from 17 patients with cystic fibrosis and from 22with other types of pulmonary problems (Table VIII).Two specimens of pus were obtained and studied in asimilar fashion. The sputa were collected in glass jarskept in the patients' home freezers at approximately- 100 C over a 2- to 5-day period and then were keptfrozen at - 200 C up to 3 weeks. Sputa collected fromhospitalized patients were frozen after a 4-hour collectionperiod.

Preparation of sputum homogenates. Individual spu-tum specimens were homogenized in distilled water witha Potter-Elvehjem homogenizer to form 10% (wetweight to volume) homogenates. The 10% homogenateswere diluted further in some experiments. In certainexperiments fresh sediments were prepared by centrif-ugation 1 of the 10% aqueous sputum homogenates at30,000 g X 15 minutes at 00 C. The supernatant fluidswere decanted, and the sediments were washed where in-dicated by resuspension to the initial volume followed byrecentrifugation.

Assay of proteolysis. Proteolysis was determinedroutinely by measuring the liberation of trichloroaceticacid (TCA) -soluble protein fragments by the Folin-Ciocalteu (F-C) test. The 10% homogenate of sputumand the reagent whose proteolysis-stimulating effect wasbeing tested were both adjusted to the appropriate pH(see below) with 0.1 N NaOH or HCl, before beingmixed in equal volumes. The mixture was then incu-bated in duplicate in a 370 C water bath. Two-ml sam-ples were removed at suitable intervals, added to 4.0 ml16% TCA, and centrifuged at 30,000 g for 15 minutes.The supernatant fractions (2.0 ml) were mixed with 4.0ml of 1 N NaOH in colorimetric cuvettes, and 1.0 mlof F-C reagent (diluted 1: 3 with distilled water) wasadded to each cuvette and mixed immediately. Exactly5 minutes later the absorbance was measured on a Cole-man Junior spectrophotometer at 675 mjA against a rea-gent blank. Proteolysis is expressed as the optical den-sity developed in a given period of incubation. A similarproteolytic assay utilizing F-C reagent was used with

1 Servall refrigerated automatic centrifuge, SM-24rotor.

1892

INDUCTION OF PROTEOLYSISIN PURULENTSPUTUMBY IODIDES

extraneous protein substrates such as casein, hemoglobin,gelatin, and albumin.

Three other methods for detecting proteolysis wereutilized to confirm the observation of iodide-induced pro-teolysis. The first of these methods assayed the tyrosinecontent of the TCA-insoluble sediment, instead of thesupernatant fluid, with the F-C reagent as follows:After centrifugation of the TCA mixture, the sedimentwas washed once with 10%o TCA, then dissolved in avolume of 1 N NaOHequal to that of the initial TCAmixture. Two ml of the dissolved sediment was thenmixed with 4.0 ml 1 N NaOH and finally with 1.0 mlof the F-C reagent; the absorbance was measured onthe spectrophotometer at 675 mgi against a reagent blank.

In a second supplementary method of assaying proteoly-sis, the absorbance of the TCA-supernatant fluid wasmeasured directly with a Beckman DU spectrophotometerat 280 mgt against a water blank.

The third supplementary method for assaying proteoly-sis involved the assay of alpha-amino nitrogen with Nin-hydrin according to the method of Moore and Stein (6).A 0.2-ml sample of the TCA-supernatant fluid wasmixed with 1.3 ml of 0.1 N NaOH (to adjust to pH 5.0)and with 2.0 ml of the Ninhydrin solution. The mixturewas heated in a boiling water bath for 20 minutes, fol-lowed by the addition of 15 ml of diluent (a 50: 50 mix-ture of n-propanol with water) and vigorous shakingfor 1 minute. The mixture was poured into colorimetertubes and the absorbance determined against a reagentblank at 570 my on a Coleman Junior spectrophotometer.

Preparation of iodinated tyrosine and thyronine solu-tionzs. The iodinated organic compounds used in this

.30- Nal

.25-

.20 NaCI

,.15-

X2 R r

15 30 60 120Minutes

FIG. 1. INFLUENCE OF VARIOUS IONS AND UREA ON

PROTEOLYSIS IN PURULENTSPUTUMAT PH 7.5 (CONCEN-TRATION OF EACHADDITIVE =1 M).

.2° .

| Nal

NaCI

Added

DNase- I nducedProteolysis I

III

15' 60 120 1 f5 ' 60 120 18030 180 30Minutes

FIG. 2. ADDITIVE EFFECTS OF 1 M NAI OR NACL ON

DNASE-INDUCED PROTEOLYSIS IN PURULENTSPUTUM.

study were difficult to dissolve in neutral buffer solutions.The following procedure was used routinely for pre-paring 25.0 ml of such solutions in 0.05 M, pH 7.5 so-dium phosphate buffer. The iodine compound wasweighed and then dissolved in 2.0 ml of 0.1 N NaOH.Fifteen ml of distilled water and 2.5 ml 0.5 M, pH 7.5sodium phosphate buffer were then added, and the mix-ture was brought to pH 7.5 with 0.1 N HCl while beingstirred constantly. The volume was adjusted to 25.0 mlwith distilled water.

Reagents. Reagent grade chemicals were used. Soy-bean-trypsin inhibitor (5X crystalline) 2 L-3,5-diiodo-thyronine and its analogues, L-3,5-diiodotyrosine, D-3,5-diiodotyrosine, L-3,5-dibromotyrosine, and tyrosine; 3

L-3,3',5-triiodothyronine, L-thyroxine, D-thyroxine, L-3-monoiodotyrosine, and DL-3,5-diiodothyronine; 4 deoxy-ribonuclease (1X crystallized) ; 5 and trypsin (2X crys-talline)2 were obtained commercially.

Results

The experimental data to be described were ob-tained from experiments with individual sputumspecimens, but each experiment was repeated withsputa obtained from at least four other individuals.The absolute proteolytic activity varied from pa-

2 Nutritional Biochemicals Corp., Cleveland, Ohio.3 Cyclo Chemical Corp., Los Angeles, Calif.4 California Corporation for Biochemical Research,

Los Angeles, Calif.5 Worthington Biochemical Corp., Freehold, N. J.

1893

IIIIIII

JACK LIEBERMAN AND NATHANIEL B. KURNICK

.30 0

C;~~~~~~~p

'A.20 /

.10 -

4.0 5.0 6.0 1.0O 8.0O 9.0

pH

FIG. 3. OPTIMAL PH CURVEFOR PROTEOLYSIS INDUCED BY1 M NAI IN A HOMOGENATEOF PURULENTSPUTUM.

tient to patient (4), but the qualitative resultsdescribed below were characteristic of all speci-mens examined.

Effect of various electrolytes

Figure 1 shows the rates of proteolysis at pH7.5 in 5% sputum homogenates in the presence ofvarious 1 Msalt solutions or urea or with distilledwater as a control. Sodium iodide (or potassiumiodide) caused a higher rate of proteolysis thanthat resulting from any of the other electrolytestested, although the other halogens, namely, chlo-ride, bromide, and fluoride, also caused some pro-teolysis. Sodium thiocyanate induced only mini-mal proteolysis. The effect of sodium cyanide or

1-

._C

4An

370 C

urea was insignificant and did not differ appreci-ably from the control.

A difference between iodide- and chloride-in-duced proteolysis was demonstrated by the ex-periment shown in Figure 2. The proteolysis re-sulting from the action of DNase (1.0 mg DNase+ 5.0 ml sputum homogenate + 5.0 ml 0.05 M so-dium phosphate buffer, pH 7.5) appeared to becomplete by the end of a 3-hour incubation period(Figure 2). At this time an equal volume of 2 MNaCl was aded to one portion of the reactant mix-ture, and 2 M Nal was added to another portionof the mixture. The addition of NaCl did not re-sult in further proteolysis, but the addition ofNaI renewed the proteolytic action.

.20

.15- Nal + Trypsin

6 _^ Nal +AddedSputum Protease

.0

NaCI + TrypsinNal Alone

15 30 60 120Minutes

FIG. 5. INFLUENCE OF 1 MNAI OR NACLON PROTEOLY-SIS BY TRYPSIN (0.2 MG PER ML) OR SPUTUMPROTEASE(0.5 ML OF UNHEATEDSPUTUMHOMOGENATEPER 5.0 ML)WITH 80' C-HEATED SPUTUMAS SUBSTRATE (0.025 M so-

DIUM PHOSPHATEBUFFER, PH 7.5).

Enzymatic nature of iodide-induced proteolysis

Hours

FIG. 4. EFFECT OF TEMPERATUREUPONPROT]

DUCED BY 1 M NAI IN A HOMOGENATEOF

SPUTUM (0.025 M SODIUM PHOSPHATEBUFFE]

Optimal pH. The proteolytic effect of sodium70° c iodide was studied at the various pH levels shown

in Figure 3. An optimum between pH 7.0 and7.5 was demonstrated. At pH values above 9.0the high electrolyte concentrations caused hydroly-sis that is nonenzymatic, as was previously ob-served with NaCl (5).

O c Effect of temperature. Proteolysis was moreactive at 370 C than at 700 or 0° C (Figure 4).The proteolytic reaction could be prevented byboiling the sputum preparation or by heating at

EOLYSIS IN- 800 C for 1 hour before mixing with sodium io-PURULENT dide (Figure 5, bottom curve). The native pro-

R, PH 7.5). teases within purulent sputum were found to be

1894

INDUCTION OF PROTEOLYSIS IN PURULENTSPUTUMBY IODIDES

completely destroyed by heating at 80' C (3).The addition of trypsin or of a small amount ofunheated sputum to the heated sputum preparationresulted in active proteolysis in the presence of1 M sodium iodide, but not with 1 M sodiumchloride.

Effect of soybean-trypsin inhibitor. Soybean-trypsin inhibitor (SBTI) has been observed toinhibit a "heat-labile" protease (inactivated at650 C) in purulent sputum when tested upon a

casein substrate, but not to inhibit a "heat-stable"protease (3). A similar result was obtained whensputum protein, instead of casein, was used as

substrate. SBTI (2.3 mg per ml) caused 70%inhibition of iodide-induced proteolysis (1-hourincubation) in an unheated sputum homogenate,but did not significantly inhibit iodide-inducedproteolysis of a 650 C-heated sputum preparation(Table I).

Optimal concentration of sodium iodide

Various concentrations of sodium iodide were

tested with a sputum homogenate to determine theeffect of iodide concentration upon the inductionof proteolysis. Two types of sputum preparationswere used for this experiment: (A) a routine spu-

tum preparation that had not been heated beforethe experiment and (B) another portion of thesame sputum homogenate that had been heated at650 C for 1 hour to destroy the "heat-labile" pro-

tease. As can be seen in Figure 6, iodide causedmore proteolysis in the unheated sputum during

.a * 20Qo

0 0.1 0.25 0.5 0.75 1 1.5 2

Sodium Iodide (Molarily)

FIG. 6. OPTIMAL NAI CONCENTRATIONFOR THE IN-

DUCTION OF PROTEOLYSIS IN (A) AN UNHEATEDSPUTUM

HOMOGENATEAND (B) A SPUTUM HOMOGENATEPREVI-

OUSLY HEATEDAT 650 C FOR 1 HOUR. The unheated spu-

tum preparation contains a "heat-labile" protease that isdestroyed by heating at 650 C for 1 hour.

a 1-hour incubation period than in the 65°C-heated preparation. Both systems demonstratedoptimal enhancement of proteolysis at 1 M Nalconcentration.

TABLE I

Effect of soybean-trypsin inhibitor (SBTI) upon the induction of proteolysisby 1 Msodium iodide in purulent sputum*

Nal-induced proteolysis (OD)tSpontaneous proteolysis

(OD)t Without SBTI With SBTI gInhi-

Sputum preparation 0 hr 1 hr AOD 0 hr 1 hr AOD 0 hr 1 hr AOD bition

Unheated sputum 0.201 0.200 0 0.237 0.483 0.248 0.210 0.280 0.068 70homogenate 0.205 0.203 0.240 0.487 0.205 0.271

650 C-heated 0.191 0.183 0 0.220 0.291 0.072 0.205 0.274 0.065 10sputum homog- 0.190 0.191 0.220 0.292 0.215 0.276enate

* SBTI powder (7.0 mg) was mixed with samples (1.5 ml) of unheated and of 65° C-heated sputum homogenates andallowed to stand at room temperature for 20 minutes. NaI (1.5 ml of 2 MNaI in 0.05 Msodium phosphate buffer, pH7.5) was added to these samples and to samples of the same sputum homogenate lacking SBTI. The mixtures were in-cubated at 370 C and assayed for proteolysis with Folin-Ciocalteu (F-C) reagent as described under methods. Spontane-ous proteolysis was measured in a mixture of 1.5 ml sputum homogenate with 1.5 ml, 0.05 M, pH 7.5 sodium phosphatebuffer.

t OD=-optical density~of simul taneousduplicate experiments.

1895

JACK LIEBERMAN AND NATHANIEL B. KURNICK

.00005 .0005 .005 .05 0.5

Molarity

FIG. 7. INFLUENCE OF ORGANIC IODIDES ON PROTEOLYSIS

IN PURULENT SPUTUM (1-HOUR INCUBATION) COMPARED

TO NAL.

Concentrations of Nal as low as 5 mMappearedcapable of inducing measurable proteolysis duringa 1-hour incubation period. However, when theincubation period was extended to 24 to 48 hours,enhanced proteolysis could be detected with con-

centrations of NaI as low as 0.05 mM(Table II).Spontaneous proteolysis was also measurable dur-ing this prolonged incubation but did not equalthat occurring in the presence of NaI.

Effect of iodinated tyrosine and thyronine com-

pounds

Various iodinated tyrosine and thyronine com-

pounds were tested for their ability to induce pro-

teolysis in purulent sputum. Figure 7 shows thelevel of proteolysis induced by these compounds at

different molar concentrations during a 1-hour in-cubation period. At 0.5 M concentration, themost active compounds, di- and triiodothyronineand monoiodotyrosine (MIT), produced approxi-mately 3 times as much proteolysis as sodium io-dide. Tetraiodothyronine (thyroxine) and di-iodotyrosine (DIT) were also more active thansodium iodide, but less active than monoiodotyro-sine or di- and triiodothyronine. Analogues ofdiiodothyronine, in which the alanine chain is re-

placed by propionic, formic, or acetic acids, were

much less effective than the thyronine compoundand showed no greater proteolysis-inducing ac-

tion than did sodium iodide. Iodoacetamide, an

inhibitor of many enzymes, promoted proteolysisin sputum similar to sodium iodide.

Tyrosine itself had no proteolysis-inducing ac-

tion, and substitution of bromide ion for iodideon the tyrosine did not allow proteolysis to pro-

ceed (Figure 8).Optimal concentration. The graphs in Figure 7

demonstrate that 0.005 to 0.0005 Mconcentrationsof the iodinated tyrosine and thyronine compoundsinduced proteolysis equal to that resulting from0.5 M sodium iodide. Further increases in con-

centration of these organic compounds promotedeven higher rates of proteolysis (Figure 7), butthis was shown to be dependent upon the concen-

tration of sputum being used as substrate (Figure9).

A series of dilutions of heat-inactivated (800 Cfor 1 hour) sputum homogenate was mixed withequal volumes of a series of DIT concentrations

BLE II

Effect of low concentrations of Nal and prolonged incubation upon proteolysis of purulent sputum*

ProteolysisSpontaneous proteolysis

5 mMNaI 0.5 mMNaI 0.05 mMNaI controlTime of

incubation OD AOD OD AOD OD AOD OD AOD

Ohr >o0.170>169 0.16>0 .164 0160>0 164 0.165>0 163

24 hr 0.365' ~ 0. 305\x 30.221\ ~230.200K,0.363/>.364 0.195 0.307>0306 0.142 0.2257223 0.059 0.20620203 0.040

48 hr 0.47°>0.472 0.303 0.413 0.251 0.310.316 0.152 0.251>0.253 009048~~~~~~~~~~~~~~~~~~~~~~~~~~~hr_.110*473/05%100.321n0 0.255/

*5% sputum homogenate in 0.025 M sodium phosphate buffer, pH 7.5.

1896

INDUCTION OF PROTEOLYSIS IN PURULENTSPUTUMBY IODIDES

in 0.05 At sodium phosphate buffer, pH 7.5.Trypsin (0.5 nml of 2.0 mg per ml in 0.001 NHCI) was added to each mixture and to DITblanks (no sputum) and incubated at 370 C for1 hour. The rate of proteolysis was determined inthe usual manner, and the control values fortrypsin autolysis in DIT were subtracted. Fig-ure 9 shows that maximal proteolysis was inducedin the more dilute sputum homogenates by lowerconcentrations of DIT than were needed to inducemaximal proteolysis in the more concentrated spu-tunm homogenates. This direct relationship be-tween DI1 and sputum concentration is sugges-tive of a stoichiometric reaction. When the con-centration of sputum homogenate was in excessrelative to DIT [i.e., homogenate concentration(wt/vol): DIT (molarity )> 1 ], the rate of thereaction was a function of DIT concentration.Similarly, when the concentration of DIT was inexcess relative to the sputum concentration, therate of the reaction was a function of the sputumconcentration. Concentrations of sputum greaterthan 5 o resulted in the same curve as that for5%o (Figure 9), indicating that substrate waspresent in excess.

Importance of the stereochemical structure ofiodinated tyrosine and thyronine. Figure 10 showsthe comparative rates of proteolysis caused bydextro- and levorotatory compounds in equimolar(0.005 M) concentrations. The mixture of DL-

0a-

CL .

L-3-Monoiodotyrosine

L-3,5-Diiodotyrosine

15 30 60 120Minutes

FIG. 8. EFFECT OF IODINATION OR BROMINATION UPONTHE PROTEOLYSIS-INDUCING ACTION OF TYROSINE (5 MM)IN PURULENTSPUTUM.

diiodothyronine and the D isomers of thyroxineand diiodotyrosine are much less active than thepure L compounds. Thus, the stereochemicalstructure of the amino acid to which the iodide isattached appears to be important for causingmaximal proteolysis within the purulent sputum.

This enhanced role for the L isomers was shown

5%sputum homogenate

_8 _O.-.we -- 1:10 dilution

I:IUU diluilon.4,,,.,.... 1:1000 dilution

. A 1:10000 dilution

0.5

Di-iodotyrosine (Molarity)

FIG. 9. EFFECT OF DIIODOTYROSINE (DIT) CONCENTRATIONON DIGES-TION OF VARIOUS DILUTIONS OF AN 800 C-HEATED SPUTUMHONIOGENATEBY TRYPSIN.

-16

0.4)

1897

I Ann A: L.&'. --

JACK LIEBERMAN AND NATHANIEL B. KURNICK

L-3,5-Di-iodothyronine

* L-Thyroxine/ / D,L-3,5-Di-iodothyronine

A.2 0 L-3,5-Di-iodotyrosine

D-3,5-Di-iodotyrosine.10 e D-Thyroxine

30 60 120

Minutes

FIG. 10. COMPARISONOF DEXTRO- AND LEVOROTATORY

ISOMERS FOR THE INDUCTION OF PROTEOLYTIC ACTIVITY IN

PURULENTSPUTUM.

to depend upon the integrity of cells in the spu-

tum preparation (Table III). When the cellscontained in the sputum were disrupted by dis-

4.0 5.0 6.0 7.0 8.0 9.0 10.0

pH

FIG. 11. OPTIMAL PH CURVE FOR PROTEOLYSIS INDUCED

BY 5 MMDIT IN A HOMOGENATEOF PURULENTSPUTUM.

solution in 2 M NaCl, or by repeatedly washingthe sediment with distilled water (which does notremove the enzyme from the sediment), or by theaddition of Triton X-100, the D and L isomers pro-

duced equal effects in promoting proteolysis inthe sputum preparation (Table III).

Optimal pH for the induction of proteolysiswith DIT. Samples of 5%o sputum homogenatecontaining 0.005 M DIT were assayed for pro

teolysis at various pH values during a 1-hour in-cubation period. Figure 11 demonstrates thatmaximal activity occurred at pH 7.5 to 8.0.

LE III

Comparison of proteolysis resulting from dextro- and levorotatory thyroxine in acellularand cell-containing sputum homogenates

Proteolysis (OD)Ratio of

With L-thyroxine With D-thyroxine proteolyticactivity

Activity Activity in L- toSputum preparation O hr 2 hr (AOD) 0 hr 2 hr (AOD) D-thyroxine

A. Cell-containing homogenates*Homogenate in 0.405 0.685 0.275 0.400 0.530 0.125 2.20

physiological saline 0.410 0.680 0.405 0.525

B. Acellular homogenatestSediment washed 5 to 6 0.428 0.745 0.314 0.417 0.720 0.296 1.06

times with distilled water 0.430 0.740 0.422 0.710(resuspended to originalvolume)

Homogenate in distilled 0.460 0.795 0.339 0.435 0.765 0.324 1.05water treated with 0.455 0.797 0.445 0.763Triton X-100

Homogenate in 2 MNaCl 0.419 0.829 0.407 0.421 0.794 0.367 1.110.422 0.825 0.428 0.788

* Microscopic examination reveals intact cells.t Microscopic examination reveals no cells, only amorphous material.

1898

INDUCTION OF PROTEOLYSISIN PURULENTSPUTUMBY IODIDES

TABLE IV

Effect of 5 mMdiiodotyrosine (DIT) upon proteolysis of various substrates at pH 7.5 by sputum protease,trypsin, and chymotrypsin*

Substrate

Caseint Denatured hemoglobins Gelatin§ Albuminl! Sputum protein¶

1% 1%0 1% 1% 5%

Proteolysis (AOD)**

No With % No With %C No With % No With % No With %Enzyme DIT DIT Change DIT DIT Change DIT DIT Change DIT DIT Change DIT DIT Change

(7j /l C7fSputum

protease, 0.265 0.237 -10.5 0.312 0.293 - 6.5 0.210 0.183 -4.3 0.040 0.015 0.001 0.106 +

10% wt/volextract in1 MNaCl

Trypsin, 0.649 0.640 - 1.4 0.316 0.387 +22.4 0.263 0.272 +3.4 0.466 0.488 + 4.7 0.000 0.175 +

0.2 mg/ml

Chymotrypsin, 0.404 0.497 +23.0 0.504 0.522 + 3.6 0.281 0.205 -6.0 0.196 0.172 -12.5 0.000 0.101 +

5,000 ArmourU/mlff

* 2.5 ml of each substrate in 0.05 Mphosphate buffer, pH 7.5 (2% wt/vol solutions, except sputum protein, 10% solution) was mixed with anequal volume of either the phosphate buffer or 0.01 MDIT in the buffer. 0.5 ml of each enzyme (type and concentration listed in the table) wasadded to each substrate mixture, and the tubes were incubated at 370 C for 1 hour. Samples (2.0 ml) were mixed with 4.0 ml of 1670 trichloro-acetic acid (TCA) at 0 hour and 1 hour. Proteolysis was assayed with the F-C reagent as described under Methods.

t Vitamin-free casein, Pfanstiehl Laboratories, Inc., Waukegan, Ill.t Hemoglobin (denatured), Nutritional Biochemicals Corp., Cleveland, Ohio.

Knox unflavored gelatin, Knox Gelatine, Inc., Johnstown, N. Y.Albumin, human (crystallized), Mann Research Laboratories, Inc., NewYork, N. Y.Sputum heated at 800 C X 1 hour at pH 8.0 to destroy intrinsic protease activity.

** ODvalues = change in optical density, average of duplicate assays.

it One-tenth the concentration of chymotrypsin was used with casein substrate.

Substrate specificity for enhanced proteolysisweith DIT

DIT was tested for its effect upon the activitiesof sputum protease, trypsin, and chymotrypsinwith substrates other than the sputum itself.Casein, denatured hemoglobin, gelatin, and humancrystalline albumin (1 % concentrations) were

used as substrates (Table IV). The data showthat 0.005 MDIT has no activating or inhibitoryeffect upon these proteases with these other sub-strates, but that DIT does allow proteolysis ofsputum protein. Wehave also observed that theautolysis of trypsin and chymotrypsin is enhancedby the presence of DIT, necessitating the inclu-sion of blanks containing only the enzyme andDIT. Appropriate blanks were also needed whenthe sputum protease was studied with other sub-strates, since the 0.5 ml of sputum homogenateused as enzyme was also readily hydrolyzed inthe presence of DIT.

Effect of DIT upon boiled sputum substrate

A 10%o sputum homogenate in distilled waterwas placed in a boiling water bath for 30 minutesand was then tested for a) the presence of a

trypsin inhibitor and b) the influence of DIT ontryptic digestion of the boiled sputum. The pur-pose of this experiment was to determine whetheror not the enhancement of proteolysis in sputumby DIT could be due to the inactivation of a pro-tease inhibitor by DIT.

One-half ml samples of trypsin (0.2 mg per ml)were mixed with 0.5-ml samples of the boiledsputum homogenate and incubated at room tem-perature for 30 minutes before the addition of 5.0ml of 1% casein (in 0.05 M sodium phosphatebuffer, pH 7.5). The mixture was incubated at

TABLE V

A test of trypsin-inhibitory activity by aboiled sputum homogenate*

Hydrolysis of casein by trypsin

In presence ofboiled-sputum In presence of

homogenate distilled waterTime of

incubation OD AOD OD AOD

Ohr 0 t 75>0. 175 01 12>O 107

15 min 0g3900>039S 0.220 0 32°0 .318 0.211

30 min 0.585> 583 0.409 0g53j5>.537 0.430

* See text for procedure.

1899

or,

1900 JACK LIEBERMANAND NATHANIEL B. KURNICK

TABLE VI

Effect of 5 mMdiiodotyrosine (DIT) upon the proteolysis of boiled sputum homogenate by trypsin*

Proteolysis of boiled sputum Trypsin autolysis Net pro-_______________________________________ _______________________________________-otelysi

A B C D of boiledWithout DIT With DIT Without DIT With DIT sputum

Time of with DITincubation OD AOD OD AOD OD AOD OD AOD (B minus D)

0 hr 0.257 0.597 0.059 0.7670.260 0.603 0.057 0.753

30 min 0.265 0.008 0.775 0.181 0.063 0.008 0.829 0.0730.269 0.786 0.068 008 0.837 0.07 oto

60 min 0.281 0.020 0.995 0.398 0.075 0.019 0.910 0.158 0.2400.277 1.000 0.078 019 0.925024

* 2.5 ml boiled 10% sputum homogenate or H20 for control + 2.5 ml 0.01 MDIT in 0.05 M, pH 7.5 sodium phosphate buffer or buffer alone+ 0.5 trypsin (2.2 mg per ml).

TABLE VII

Assay of iodide-induced proteolysis by various techniques*

Method of assay

Folin-Ciocalteu reagent Absorbance at 280 mgo Ninhydrin reagent

TCA-supernate TCA-sediment TCA-supernate TCA-supernate

OD AOD OD AOD OD AOD OD AOD

A. Spontaneous proteolysis0 hr 0.155 0.550 0.252 0.500

0.153 0.551 0.250 0.505

30 min 0.150 0 0.550 0 0.243 0 0.490 00.155 0.555 0.251 0.500

60 min 0.158 +0.004 0.565 0.005 0.247 0 0.485 0

120mm ~~~~~0.160055 .4 0.451120 min 0.16600 +0.006 0.543 -0.008 0.250 +0.001 0.500 +0.0020.160 0.540 0.253 +0.002

B. Nal (1 M)-inducedproteolysis

0 hr 0.210t 0.470 1.431t 0.5150.205 0.475 1.426 0.515

30 min 0.250 +0.044 0.421 -0 050 1.468 +0.041 0.530 +0.0150.253 0.424 1.471053

60 min 0.327 +0.116 0.3857 -0.087 1.509 +0.084 0.580 +0.063

120 min °04329° +0.222 0.295 -0.180 1.596 +0.164 0.640 +0.1430.429 0.2~~91 1.590 +014 0.675

C. DIT (0.005 M)-inducedproteolysis

0 hr 0.573§ 0.361 1.461§ 0.5200.610 0.365 1.498 0.560

30 min 0.679 +05 0.305 -00 1.620 0.5800.675 +0.085 0.300 0.060 1.598 +0.120 0.580 +0.040

60 min 00.725 +0.129 0.243 -0.121 1.728 +0.250 0.610 +0.075

120 min 0.820 +0.249 0.157 -0.210 2.004 +0.053 0.680 +0.1500.861 0.150 1.961 0.700 +.5

* Samples of a 10%, purulent sputum homogenate were mixed with equal volumes of 0.05 M, pH 7.5 sodium phosphate buffer, or 2 MNaI in thesame buffer, or 0.01 MDIT in the buffer, and incubated at 370 C. The mixtures were sampled at intervals and mixed with 2 parts 16% TCA.The TCA-supernatant fluids and sediments were then subjected to the various assays as described under methods.

t The slightly higher TCA-supernatant fluid ODand lower TCA-sediment OD, as compared to that seen with spontaneous proteolysis, is dueto the immediate nonenzymatic release of a small amount of acid-soluble material by the NaI.

* Sodium iodide and TCA react to form a compound that absorbs at 280 mjs and is associated with a yellow discoloration of the solution. Ablank containing only TCAand NaI gives an absorbance of slightly greater than 1.0 OD, but the absorbance increases with time. For this reasonthe measurement of absorbance at 280 mix is a poor method for measuring proteolysis in the presence of Nal. The reaction of Nal with TCAdidnot affect the Folin-Ciocalteu (F-C) color development, however.

§ The initial high ODis due to the added DIT and to the immediate nonenzymatic release of a small amount of acid-soluble material by theDIT. This effect is apparently more marked on the absorbance at 280 m/A than on the intensity of the F-C reaction, because of the greater slopeof the absorbance: concentration curve at 280 miA.

INDUCTION OF PROTEOLYSISIN PURULENTSPUTUMBY IODIDES

370 C and sampled at 15- and 30-minute inter-vals. Comparison to a similar preparation uti-lizing distilled water instead of boiled sputumhomogenate revealed that the sputum did not in-hibit trypsin activity (Table V).

When the boiled sputum homogenate was usedas substrate for trypsin (0.2 mg per ml finalconcentration), it resisted proteolysis completelyin the absence of DIT but was hydrolyzed bytrypsin in the presence of 0.005 M DIT (TableVI).

Since trypsin-inhibitory activity could not be

demonstrated in the boiled sputum preparation,and since DIT can induce tryptic proteolysis ofthe boiled sputum, this action of DIT probablyis not dependent upon the removal of a proteaseinhibitor.

Other methods of assay

Data obtained from four different methods forassaying proteolysis are shown in Table VII.All four methods showed minimal spontaneousproteolysis in a 10% homogenate of purulent spu-tum during a 2-hour incubation, but all the assay

TABLE VIII

A survey of DIT-induced proteolysis in purulent sputum and pus

Units 1-hr 1-hrATEEase spontaneous DIT-induced

Age Sex Diagnosis activity* proteolysis proteolysis

37 M Asthma46 M Pneumonia69 M Asthma65 M Emphysema47 M Chronic bronchitis43 M Chronic bronchitis35 F Bronchitis71 M Tuberculosis46 M Bronchogenic carcinoma43 M Bronchitis43 M Bronchiectasis57 M Lung abscess74 M Inactive tuberculosis39 M Bronchiectasis46 M Inactive tuberculosis40 M Asthma

4 M Cystic fibrosis35 M Asthma

8 M Cystic fibrosis68 M Emphysema13 F Cystic fibrosis27 M Pneumonia76 M Bronchogenic carcinoma76 M Asthma13 M Cystic fibrosis6 M Cystic fibrosis

11 M Cystic fibrosis7 M Cystic fibrosis

73 M Emphysema54 F Bronchiectasis14 M Cystic fibrosis73 M Bronchiectasis10 M Cystic fibrosis

F Pus from right kidney abscesst16 F Cystic fibrosis

F Pus (empyema)t16 M Bronchitis12 M Cystic fibrosis

9 F Cystic fibrosis7 F Cystic fibrosis

13 M Cystic fibrosis9 F Cystic fibrosis

10 F Cystic fibrosis15 F Cystic fibrosis

0.250.50.51.01.11.251.62.02.12.64.05.65.96.07.59.8

12.515.017.518.020.020.622.522.523.026.326.328.129.032.040.042.548.050.051.356.050.071.075.079.080.682.085.0

100.0

0.0030.0020.0060.0040.0020.0050.0080.0030.0020.0000.0020.0060.0020.0040.0000.0000.0000.0000.0000.0040.0000.0020.0030.0060.0060.0030.0000.0010.0000.0010.0000.0020.0040.0010.0060.0050.0000.0030.0000.0000.0000.0000.0000.004

0.0360.0410.0610.1320.1180.0790.0500.1560.1950.2210.2100.1150.1180.2570.1400.1380.1320.2440.2570.2020.2740.2580.2060.0780.1720.2450.2760.2940.0790.2220.2570.2350.1790.0700.2280.1040.1420.2550.2890.2490.2440.2580.2850.294

* ATEE = N-acetyl-L-tyrosine ethyl ester.t Prepared as 10% homogenates in distilled water.

1901

JACK LIEBERMANAND NATHANIEL B. KURNICK

techniques demonstrated that active proteolysistook place in the presence of 1 MNal or 0.005 MDIT.

Reproducibility of the DIT effect on proteolysisin purulent sputa

Forty-two specimens of purulent sputum andtwo samples of pus were prepared as 10% ho-mogenates in distilled water. Each specimen wasassayed with 0.01 MN-acetyl-L-tyrosine ethyl ester(ATEE) substrate as an estimate of proteasecontent (4). Samples (2.5 ml) of each homoge-nate were then mixed with equal volumes of0.05 M sodium phosphate buffer, pH 7.5, tomeasure spontaneous proteolysis, or with 0.01 MDIT in the buffer at pH 7.5, and incubated at370 C. At intervals of 0, 30, 60, and 120 minutes,1.0-ml samples were removed and mixed with 2.0ml 16% TCA. The TCA-supernatant fluid wasthen tested for the degree of proteolysis with theF-C reagent.

DIT-induced proteolysis was linear with time.Table VIII contains the data for spontaneousproteolysis as compared with DIT-induced pro-teolysis for the 1-hour incubation period. Thespecimens are listed in the order of increasingactivity with ATEE substrate. None of the speci-mens showed any appreciable spontaneous proteol-ysis during the 1-hour incubation, but all of themshowed enhanced proteolysis in the presence of0.005 M DIT. The DIT-induced proteolysiswas roughly proportional in rate to the hydrolysisof ATEE by the sputum specimens, up to anATEE hydrolytic activity corresponding to 25 U.Sputa that contained higher levels of ATEE hy-drolytic activity did not demonstrate furtherincreases in DIT-induced proteolysis becauseof the limiting concentration of DIT (Figure 9).

Discussion

In the past it has been thought that iodides actto thin purulent respiratory secretions simply byincreasing the volume of aqueous secretion fromthe bronchial glands. However, our experimentssuggest that iodides may have a more specific roleand that they liquefy purulent secretions by in-ducing the enzymatic hydrolysis of protein. Thisaction of iodides requires the presence of a pro-teolytic enzyme: either inherent leukocytic pro-

teases or an added protease such as trypsin.Since the reaction is enzymatic, inhibition of theenzyme prevents proteolysis.

Although the data demonstrate optimal proteol-ysis at a concentration of 1 M sodium iodide,considerable proteolysis was apparent at 0.005 M(Table II and Figure 7). Even lower concentra-tions may be clinically effective, since the amountof proteolysis that must take place to reduce theviscosity of sputum is very slight compared to thelevels of proteolysis attained under optimal itvitro conditions. For example, we have demon-strated effective liquefaction of whole purulentsputum by the addition of trypsin (7), althoughthe level of measurable proteolysis resulting fromtryptic activity is minimal.

The mechanism whereby iodides induce pro-teolysis does not appear to depend upon the dis-sociation of nucleoprotein complexes and thus dif-fers from the manner in which deoxyribonucleaseand high NaCl concentrations act. The reactionis substrate specific in that proteolysis of someprotein substrates other than sputum, such ascasein, albumin, hemoglobin, and gelatin, was notaffected by the iodides. In addition, this effect ofiodides is not enzyme specific, since the action oftrypsin and chymotrypsin upon a heated sputumpreparation (lacking inherent protease activity)was also enhanced by the iodides. The iodidesappear to act by altering the potential proteinsubstrate in sputum so that it is more readily hy-drolyzed by proteolytic enzymes. The nature ofthis protein substrate and the changes caused bythe iodides is under study. The splitting of glyco-proteins, analogous to the splitting of nucleopro-teins by DNase, is suggested to be the mechanismwhereby iodides induce proteolysis.

The naturally occurring iodinated thyroidalcompounds were much more effective for initiatingproteolysis in a homogenate of purulent sputumthan were the inorganic iodides. Lower concen-trations of these natural compounds were required,and faster rates of proteolysis resulted. It ispossible that the reaction is specific for iodinatedtyrosine or thyronine and that inorganic iodidemust first form iodinated amino acids to have aneffect. In the absence of iodination, on the otherhand, the parent amino acid (tyrosine) lackedthis activity, as did brominated tyrosine. Theaction of the organic iodine compounds appeared

1902

INDUCTION OF PROTEOLYSIS IN PURULENTSPUTUMBY lODIDES

to result from a stoichiometric reaction betweenthe compound and sputum protein, with either theorganic iodine compound or the sputum proteinbeing the limiting factor in determining the rateof proteolysis when the other was present inexcess.

Use of both dextro- and levoisomers of the io-dinated thyronine and tyrosine compounds sug-gests that some of the substrate affected by theiodides is intracellular. The levorotatory com-pounds were consistently more effective than thedextrorotatory compounds, but these differenceswere eliminated by disrupting the cells containedin the sputum preparation.

These observations raise the question of whetherthyroxine and its precursors directly affect pro-teolysis in vivo in cellular metabolism. Evidencethat thyroid hormone influences enzymatic ac-tivity has long been sought, but very few suchreactions have been found to be affected by thy-roxine in vitro (8, 9). Barker has shown suchan effect in carbohydrate oxidation (10), but theeffect lacked specificity since it could be producedby metabolically inactive analogues of thyroxine.The test system employed by Barker, however,was essentially a cell-free system, and possiblyini vivo specificity of thyroid hormone dependsupon its ability to cross the cell membrane. Inour experiments no difference between the D andL compounds was observed when a cell-free sys-tem was used, but a definite advantage of themetabolically active L compounds over the D iso-mers was apparent when cells were contained inthe system. However, the reaction is not com-pletely specific for metabolically active compounds,since iodinated tyrosines (MIT and DIT), thoughtto be metabolically inactive, similarly enhanceproteolysis in sputum. Other metabolically inac-tive analogues of thyroxine (diiodothyropropionic,-acetic, and -formic acids) are much less activethan iodinated thyronine and tyrosine, suggestingsome selective advantage for the thyroidal hor-mones. The order of effectiveness of the organiciodides for inducing proteolysis (diiodothyronine

triiodothyronine > monoiodotyrosine > thyrox-ine = diiodotyrosine) is not correlated with themetabolic activity li vivo. The difference betweenthe metabolic effect in vivo anid the proteolytic-ellhancing effect in vitro 1may (depend upon dif-

ferences in penetration of viable cells by the io-dinated compounds.

The lowest concentration of DIT or L-thyroxinethat enhanced proteolysis measurably in our sys-tem was 5 uM. This is approximately two ordershigher than is likely to occur in tissues. How-ever, factors of time and local concentrations (ofhormone, proteolytic enzyme, and of specific sen-sitive substrates) may provide suitable conditionsfor physiologic effects.

As a result of these studies, certain historicalobservations concerning the use of iodides appearto have a scientific basis. The use of iodides fordissolving syphilitic gummas, and the fear of us-ing iodides in tuberculosis because of the possi-bility of breaking down tubercles and spreadingthe infection, could both be related to the pro-teolytic action resulting from iodides in exudates.In 1915, Jobling and Petersen (11) investigatedthe action of iodides on necrotic material and con-cluded that iodides acted to soften tubercles andsyphilitic gummas by causing a general reductionin antitryptic activity, thereby allowing the pro-teolytic breakdown of the caseous material. Theseearly studies closely parallel our own, but wefound no evidence for the implication of a pro-tease inhibitor. Indeed, we have demonstratedthat boiled sputum lacks an inhibitor of trypsin,but nevertheless its proteolysis by trypsin is en-hanced in the presence of organic iodides. Ourexperiments suggest that the iodides interact withthe substrate to produce enhanced susceptibility toproteolytic enzymes.

The highly effective action of iodinated-tyrosinecompounds upon the induction of proteolysiswithin purulent sputum (or other purulent ma-terial) suggests that such compounds may becomeimportant therapeutic adjuncts in the treatmentof chronic respiratory disease by aiding in theliquefaction of viscid respiratory secretions. Wehave observed that whole sputum from patientswith cystic fibrosis is liquefied when mixed withDIT in vitro, indicating that a reduction in vis-cosity accompanies proteolysis. Compounds suchas monoiodotyrosine and diiodotyrosine, in con-

trast to the iodinated thyronines, have no sys-temic calorigenic effect and may be found usefulfor thinning respiratory secretions without caus-

ing calorigenic side effects. Tlhe clinical effective-

1903

JACK LIEBERMAN AND NATHANIEL B. KURNICK

ness of such compounds, when administered byaerosol, is now being investigated.

SummaryIodides were found to induce proteolysis in

purulent sputum from patients with cystic fibro-sis and other forms of respiratory disease. Pro-teolysis was shown to be enzymatic in nature andto depend upon the presence of natural leukocyticproteases or an added protease such as trypsin.The reaction was inhibited by conditions thatinhibited the proteolytic enzyme. Naturally oc-curring iodinated thyroidal compounds were muchmore effective for inducing proteolysis than werethe inorganic iodides. The L-isomers of thesethyroidal compounds promoted a higher rate ofproteolysis than did the D-isomers, when intactcells were present in the sputum preparation.

The question is raised as to whether thyroxineand its precursors affect proteolysis in vivo in cel-lular metabolism. The possible clinical use ofiodinated tyrosine compounds is suggested forthinning viscid purulent respiratory secretions.

AcknowledgmentWe are indebted to Mrs. Betty M. Trimmer for her

skilled technical assistance.

References1. Goodman, L., and A. Gilman. The Pharmacological

Basis of Therapeutics, 2nd ed. New York, Mac-millan, 1955, p. 817.

2. Opie, E. L. The enzymes in phagocytic cells of in-flammatory exudates. J. exp. Med. 1906, 8, 410.

3. Lieberman, J., and N. B. Kurnick. Characterizationof the proteolytic activities in purulent sputum.Submitted for publication.

4. Lieberman, J., and N. B. Kurnick. Proteolytic en-zyme activity and the role of deoxyribonucleic acidin cystic fibrosis sputum. Pediatrics 1963, 31,1028.

5. Lieberman, J., and N. B. Kurnick. Influence of de-oxyribonucleic acid content on the proteolysis ofsputum and pus. Nature (Lond.) 1962, 196, 988.

6. Moore, S., and W. H. Stein. Photometric Ninhydrinmethod for use in the chromatography of aminoacids. J. biol. Chem. 1948, 176, 367.

7. Lieberman, J. Enzymatic dissolution of pulmonarysecretions. Amer. J. Dis. Child. 1962, 104, 342.

8. Barker, S. B. Biochemistry in The Thyroid, 2nd ed.,S. C. Werner, Ed. New York, Harper & Row,1962, p. 102.

9. Barker, S. B. Mechanism of action of the thyroidhormone. Physiol. Rev. 1951, 31, 205.

10. Barker, S. B. In vitro action of thyroxine analogson succinate and malate oxidation. Endocrinology1957, 61, 534.

11. Jobling, J. W., and W. Petersen. The therapeuticaction of iodin. Arch. intern. Med. 1915,-15, 286.

1904